The IAU famously decided that Pluto is not a planet any more, calling it a "dwarf planet" instead, along with Ceres. However, their definition works as well as it does only because of a historical accident of our solar system and the bodies we have discovered in it so far. There are good reasons for thinking it has a "use before" date too, as we'll see. We don't know when that date is, but we will probably reach a point where we can't use it any more, at some point in the next two or three decades. The "use before "date may well be sooner than that. So why not change right away?

So, what's the alternative? It's the "geophysical definition" already used by many planetary scientists, according to which all bodies large enough to reach equilibrium under gravity are planets. That includes  Ceres too and all the other "dwarf planets" like Haumea, and Eris. It also includes any moons large enough to be in equilibrium, as "moon planets". That way we also get to call our Moon a planet! It's the same for Europa. Ganymede, Callisto, Io, Titan, Triton, Enceladus, .... Moon planets. I think that makes a lot of sense, as our discoveries lead us to see these places more and more as complex small worlds in their own right. Our solar system is rich with diverse miniature worlds, planets in their own right, and some of the most interesting ones happen to be orbiting other larger planets.

Some astronomers don't have strong views one way or another. Emily Lakdawalla, a planetary geologist and one of the senior scientists at is quoted as saying:

“When people ask me about my opinion on Pluto’s planethood, my opinion is that it’s the least interesting conversation to have about Pluto. I would so much rather talk about the exciting things that we’ve seen on Pluto … How did it form? How did its moon form? What does it have to say about the formation of our solar system? Is it like the other worlds in the Kuiper Belt, or not? … So what word a few English-speaking people use to talk about it is really the least interesting conversation I can think of.”

I'd agree with her that it's hardly the most important thing about Pluto. However, the IAU cared enough about it to change the definition. So, I think it's reasonable for some astronomers to care enough about it to change the definition a second time. 

So what are some reasons why it might matter? For one thing, whether you call something a planet influences how we think about it. Would the general public be more interested in the Moon and Europa and Ceres if we called them all planets? I think they might well. It helps convey how interesting and complex we now know them to be, and how large they are too, most of them at least country sized in surface area, and many, continent sized. 

Also when the basic terms you use are confusing, then it can lead to confused thinking. I trained as a mathematician and we put a lot of care into choosing the clearest possible terms and definitions in a theory. Probably nobody is going to actually get lead into false conclusions and mistaken proofs as a result of the IAU definition, but the clearer our language, the easier it is to talk about things. It's something that happens to interest me, and if perhaps it's one of the least interesting conversations in planetary science, I think it has some interest all the same. 

Also as we look at this question, then along the way we come across some fun ideas about present and future discoveries. When you look at the things we may find out in our solar system and about exoplanets which may challenge the IAU definition in the next few decades if not sooner, those may be very interesting indeed


The IAU definition only seems precise because our system happens to have a big gap between the smaller objects Pluto, Ceres etc that are not orbit clearing, and larger objects like Mercury and Mars which are. Proponents who say it is a much more precise definition than the geophysical one can only say this because of this gap.

  • I will show that this is just a historical accident. There is nothing in theories of planetary formation to prevent a slightly larger than Ceres sized object in the place of Mercury, which would be orbit clearing, and a Pluto sized object in place of Mars, or Ceres, which would not be orbit clearing. Or any number of other objects, close to the line, some small and orbit clearing, others large and not, and others that are right on the line.

We just happen to live in a solar system that has masses and orbits of the main objects just right to make the IAU definition work for a while. 

  • Mike Brown predicts that we are likely to find Mars sized objects in the outer solar system that fail to be orbit clearing in the near future. It would just take one such discovery to "cause astronomers to go into a tizzy again" as he says about this idea.

So - the IAU definition does work right now, for those who have a strong interest in dynamics and are less interested in the makeup of the objects. 

It is a less comfortable definition for planetary geologists because they are more interested in what an object is like than how it interacts gravitationally with its neighbours in the solar system.  Some of them at least just call Pluto and these other rounded objects planets and ignore the IAU definition, even in published papers and books.

  •  I think it has a use before date. It could happen some time in the next few decades, maybe even less than a decade, when we find Mars sized or larger objects in the outer solar system.

    Or else, it might happen when we find a Juno sized non rounded asteroid orbiting a red dwarf that is still orbit clearing. Or, a distant gas giant that is not orbit clearing orbiting another star. Or even, even a sufficiently distant non orbit clearing gas giant in our own solar system.

  •  It is no more precise than the dwarf planet definition. We could easily be in a situation where we find a distant Earth sized object out there and we don't know if it is a planet or not according to the IAU definition because our observations don't constrain its mass well enough to tell. Also the different discriminants can lead to different conclusions too. We are used to vague things like that in astronomy, for instance the boundary between a brown dwarf and a planet is like that too. We often don't know if an object is a brown dwarf or a planet  

I will briefly describe how we got into this situation. But I don't want to get enmeshed in discussing the politics of how it happened, lots of people say different things. The main thing is to look forward to the future. 

  • Irrespective of how much support the IAU definition has , or why some people are so keen on it as a way of defining the word "planet" - can it continue to be used into the future as a definition of this word?

I think myself that if we found a Mars sized object in the outer solar system then everyone except a few diehard astronomers would just call it a planet. Can you imagine a TV presenter calling it a "dwarf planet"? And going on to explain that it is not a planet? Exoplanet discoveries could overturn the IAU definition perhaps rather more slowly. But one way or another I'd be very surprised if we still use the IAU definition 50 years from now and it could easily collapse as a viable definition of a planet as soon as a decade from now or less.

So most of this article is about that. Some of it also covers the question of what exactly should be covered by the geophysical definition? Should it include contact binary planets if such exist and are such possible? I argue that it should, though in its current form it does not. If you are a mathematician or astronomer, please bear with me as I explain the calculations to those who are not. Thanks!

Let's start with the historical background.


The IAU definition requires an object to be able to "clear its orbit". For instance, Earth clears its orbits of asteroids and comets within around twenty million years. On that time frame nearly all the asteroids and comets that do close flybys of us will be deflected to hit the Sun, Jupiter, or one of the other planets or moons, or will be ejected from the solar system. Since Earth clears its orbit, it counts as a planet.

Whether an object is "orbit clearing" depends not just on its size, but how far it is from the Sun. As it turns out, a Mercury sized object, if it was positioned a little beyond Neptune would no longer count as a planet according to the IAU. If you position Mars or Earth far enough from our Sun, they also would cease to be planets. After that, Neptune and eventually, larger gas giants like Saturn or Jupiter, if far enough from the sun, are no longer planets according to their definition.

Pluto and Charon placed in front of the Earth, image by NASA. When Mike Brown and others found Eris, which at the time was thought to be slightly larger than Pluto then the question became very acute "Is Pluto a planet?". The IAU said no, and Mike Brown, discoverer of Eris, said "Pluto is Dead" after their announcement.

It made sense at the time of the IAU decision, not to all astronomers by any means, but to those most interested in dynamics of solar systems rather than planetary geology. Ceres, diameter 950 km, at the time was called an asteroid as it still is, and is rounded into a spherical shape. Pluto, only a little over twice its diameter, at 2,374 km was called a planet, as was Mercury, about twice the diameter of Pluto at 4,879 km. Astronomers were finding more and more bodies approaching the size of Pluto and eventually Eris which seemed to be larger than Pluto. Until the discovery of Eris then we were using an informal definition that Pluto was a planet but none of the smaller bodies were. But with Eris, then the question was "Do we call Eris a planet? If not, why not?".

Eris is now known to be very slightly smaller than Pluto (Eris: 2,326 ± 12 kilometers (1,445.3 ± 7.5 mi), Pluto, 2,372 ± 4 kilometers (1,473.9 ± 2.5 mi) ). But it's likely we find other bodies like Eris slightly larger than Pluto eventually. So the question was, do we call Eris a planet, or Pluto a non planet. If Eris is a planet, what about bodies slightly smaller than Pluto, and where do you draw the line?

The dynamics showed a natural dividing line between Pluto and the next largest planet Mercury based on this property of orbit clearing. So that is what they went for. 


However, could we continue to use this definition, if we find bodies like Mercury, Mars or larger too far away from the Sun to “clear their orbits” and so, too far away to count as planets? Some astronomers think it's not only possible that we find bodies like that, but even likely that we do. That includes Mike Brown, discoverer of Eris and known for his support of the IAU definition, who announced "Pluto is dead" after the IAU meeting. Yet he also thinks we are likely to find these non orbit clearing Mars sized bodies or larger in the outer solar system eventually, as we will see.

If we don't find such planets in our own solar system, we are bound to find them sooner or later in other star systems and that would make it very awkward terminology.

So, this is not for nostalgic reasons. We can't roll the solar system back to the way it seemed to be when everything seemed simple and Pluto was a planet. It's because it's a natural distinction to make, and to many planetary scientists, this is the more natural way to go forward after the discovery of Eris. 

It's a natural distinction because something large enough to be rounded under gravity is normally going to have geological features too, differentiated interior and a large surface area. The smallest ones, perhaps around 200 kilometers in diameter, still have a surface area similar to that of England, or Arizona.

Instead of restricting the word planet, we should make it more inclusive. We have plenty of stars, asteroids and comets. Why not have dozens and eventually hundreds of planets in our solar system too? One way or another I can't see the existing definition working as soon as we have orbit clearing planets that are smaller than non orbit clearing "non planets". So why not update it right now? The IAU isn't likely to do that, but we can, in our ordinary conversations and in articles and books. Planetary scientists already do this in their publications.

As someone who is keen on sending humans to the Moon along with robotic explorers to our nearby "Eighth continent" larger than Africa, I'm particularly partial to the idea of calling it a "Moon planet". And the same also for Europa and even Enceladus. It helps us to understand how varied and complex they can be. Language can be quite powerful in its effect. Would the general public have lost interest in the Moon so quickly in the 1970s if it had been called a "Moon planet"?

So anyway let's look at this in more detail.


The IAU (International Astronomical Union) made the decision, but immediately afterwards there was a fair bit of criticism of how the decision was made by other astronomers, especially planetologists.

For instance, only 5% of the members of the IAU were present to vote, and only a bit over 15% of the 2700 scientists who attended the conference. That's partly because the vote was held on the last day of the conference in Prague when many of the participants had already left or were getting ready to fly out from Prague.

Also, it’s no longer “de rigueur” for a working astronomer to join the IAU. In total 420 astronomers voted out of a worldwide group of tens of thousands of professional astronomers. Those may represent as few as 1% of the professional astronomers world wide.

So, were they representative enough of the general astronomical community to make this decision? Should non attending members been involved as well, and what about astronomers not in the IAU? Was the poll wide enough in its scope?

Inside the planet definition process

Though the decision got a lot of publicity it was made by only 333 members (the ones in favour of the decision). A petition to the IAU presented five days later had 300 signatories. It said

“We, as planetary scientists and astronomers, do not agree with the IAU's definition of a planet, nor will we use it. A better definition is needed.”

Alan Sterns commented: “From the number of signatories that the petition received in a few days, it’s clear that there is significant unhappiness among scientists with the IAU’s planet definition, and that it will not be universally adopted by scientists and text book writers. To achieve a good planet definition that achieves scientific consensus will require more work.”

For the complete list of signatories, and more about this petition, see Laurel Kornfeld’s post: 87 Years of Pluto: Complete List of Signatories Who Signed 2006 Petition Rejecting IAU Definition

And indeed planetary scientists and other astronomers do continue to use the word “planet” to refer to Pluto in their peer reviewed papers. The IAU decision is not binding or a prerequisite on publication.

For instance here is a paper published in 2012, six years later, without a single occurrence of the word “dwarf” in it. The paper just talks about the “planet Pluto” throughout: Speckle Camera Imaging of the Planet Pluto published in the peer reviewed journal, the Publications of the Astronomical Society of the Pacific.

Also here is a text book published in 2012 that refers to it just as the ”planet Pluto” without qualification.

So, you don’t have to call Pluto a dwarf planet even in a text book or peer reviewed paper, as those examples show. If you want to call Pluto a planet, go ahead!

However what I say here is irrespective of the number of astronomers involved in the decision. A decision could be flawed even if it is a majority view of astronomers at the time.

I happen to think that the decision is entirely dependent on accidents of our current knowledge.

I will aim to show here that we could easily reach a situation in the near future where it becomes completely untenable as a definition to almost anyone.

In short, I think this was a decision with a “use before date”. We just don’t know what that date is yet, but probably some time in the next couple of decades or so.


So, let’s look at this in more detail. First I wish to show how self gravitating bodies like Ceres and Pluto fit in naturally with the usual everyday idea of planets as other worlds like our own.

If we use the natural definition based on whether it self gravitates then you just need to look at it to see if it is a planet or not.

What does it matter if it is smaller than we are used to for planets, like Pluto? Anything big enough to be rounded like that is likely to have an interesting and varied geology and is bound to have large areas of its surface to explore.

Vesta is just on the verge of becoming completely rounded by gravity, so would be a borderline planet.

4 Vesta - Wikipedia

Vesta’s area of 800,000 square km is similar to Pakistan and to New Zealand. It’s a little too small to self gravitate, but it does have a differentiated core.

Other borderline planets would be the tiny ringed minor planets Chiron and Chariklo.

Artist’s impression of Chariklo. Chiron is similar.

Chiron is just over 200 km in diameter so would have an area similar to England or Arizona.

The next step up is to bodies like Ceres:

Ceres has a surface area of 2.77 million square km, similar to India or Argentina. It’s large enough to count as a complex little world of its own. Although it’s only 36% the size of Australia, I think if we had a large island that size in our seas somewhere, isolated like Australia, it might well count as a continent rather than an island. Whether or not, it’s a pretty huge area.

Why not call it a planet?

As for the Pluto system, well, we would call both Pluto and Charon planets, so it would be a dual planet system, with four other smaller non planet moons. Pluto is slightly larger than Russia (area 17.1 million km²), the largest country in the world, at around 17.6758 million km² based on the new diameter of Pluto of around 2,372 km. It's more than twice the area of Australia (7.692 million km²), nearly twice the area of the USA (9.834 million km²), and it's also larger than Antarctica (14 million km²)

Here is what New Horizon’s trip was like:

What would it be like to land on Pluto :).

Video info: “What would it be like to actually land on Pluto? This movie was made from more than 100 images taken by NASA’s New Horizons spacecraft over six weeks of approach and close flyby in the summer of 2015. “

“The video offers a trip down onto the surface of Pluto -- starting with a distant view of Pluto and its largest moon, Charon -- and leading up to an eventual ride in for a "landing" on the shoreline of Pluto's informally named Sputnik Planitia. “

It’s a colourized version of the original black and white “Imagine a landing on Pluto”

Here is a flyover in 3D generated from the images, the smallest features are 1 km across:

It’s mountains are 11,000 foot high, as high as the Rockies. And it’s active and crater free. Water ice - and also, nitrogen. This is a news report from the flyby itself, so they have learnt a lot since then, but it gives a good overview and it helps recapture the excitement of those first images of Pluto and Charon:

How can you not call this a planet!


At the moment we happen to be in a solar system where all the bodies in gravitational equilibrium that are as large as Mercury and larger clear their orbits and the ones as small as Ceres and smaller can’t.

But - it just takes one discovery of a Mercury sized or larger “non planet” beyond Neptune and that could turn that on its head. The proposed “planet 9” is just on the cusp between a planet and a non planet. It is marginally a planet according to the IAU definition if it is exactly as predicted, though closer to the rather arbitrary dividing line than any other planet to date.

But according to some ideas it could be not just one body but several smaller bodies. If so, they could easily be non planets, not clearing each other from their orbits. There are also other ideas for instance a planet 5 times the mass of Earth at 100 au.

The vertical axis shows its mass in Earth masses. Full size image here. This graph is adapted from Ethan Siegel's Figure 1 in his  article in Forbes magazine: "The Science Has Spoken: Pluto Will Never Be A Planet Again" - I've extended it to beyond one light year to show how Jupiter would be a "dwarf planet" if it orbited that far from the Sun.

As you can see from that diagram, Mercury, if moved to a bit beyond Neptune, would be a dwarf planet non planet according to the IAU.

An Earth sized body at the distance of Planet 9’s proposed orbit also would be non orbit clearing

We could have a Jupiter sized dwarf planet non planet over a light year from the Sun.

We could have Saturn and even Jupiter sized non planets in the distant Oort cloud, with the Jupiter orbiting at one light year or more away - it’s possible.

You might wonder how we could have a planet so far away, a quarter of the way to the nearest star - but stars rarely pass close to each other and disturb each other’s planets less than you might think. This is the same orbit as was proposed for the Nemesis and Tyche “planet X” candidates.

We are now pretty sure there isn’t even a brown dwarf there now as a result of the WISE infrared survey, but as for a large gas giant, that’s not ruled out at all yet.

I think that this situation is just very very confusing.

The idea of equating “dwarf” with “clears its orbit” has nothing to do with the way “dwarf” is used in English so that part of the IAU definition I think just doesn’t work in the English language. Orbit clearing is an important concept, it’s just the idea that “non dwarf” = “orbit clearing that I see as the issue here.

Also another thing that’s confusing about the definition is the idea that the adjective “dwarf” applied to “planet” makes it a “non planet”. Since when was a dwarf x not an x? Is a dwarf person not a person?

Note that planetary scientists generally don’t like the idea and just continue to call Pluto a planet, ignoring the IAU. The definition is mainly promoted by people who don’t study planets, especially planetary geology.

Geologically then any planet with a differentiated interior has a lot in common with other planets of the same type and when they are rounded by gravity generally they also have a differentiated interior. You could have a further distinction between planets with a differentiated interior and others that are homogenous, which would then add in Vesta as a planet too, but it’s not so easy to tell just by looking at it, whether it has a differentiated core.

There isn’t any other physical difference to distinguish the IAU planets from the "dwarf planets". For instance we can’t use continental drift - if we did, only Earth would count as a planet. We can’t use presence of active volcanoes as Mercury wouldn’t count as a planet.

Basically the IAU definition is orientated towards people who are much more interested in the effect of bodies on their surroundings than on the bodies themselves, even their size. It works so long as the definition aligns to some extent with the general public’s intuitive idea that a planet is a “big thing” and smaller bodies are not planets.

But it would just take one discovery, which could happen tomorrow, of a planet larger than Mercury beyond Neptune - and that would be the end of it, I am pretty sure. I just can’t see the general public accepting the idea of a non planet beyond Neptune larger than the planet Mercury.


Even worse, we could easily discover a non orbit clearing gas giant. We would then have to call it all of these at once, according to the IAU definition:

  • “Gas giant”
  • “Dwarf planet”
  • “Non planet”

It would be a “dwarf planet gas giant non planet”!

I don’t think the definition will last too long once we have mouthfuls like that to cope with.

As I said in the introduction, I’m a mathematician by training and one of the most important things in maths is to start with clear concepts. I can’t imagine a mathematical theory that uses language like this as mathematical terms being anything less than very confusing. This has to be confusing to astronomers too if not in such an immediately obvious and acute way as it would in a mathematical theory.


The IAU definition did get one thing right, when they required a planet to be in hydrostatic equilibrium. Here are the two competing definitions:

The IAU's original, 2006 definition:

  1. It needs to be in hydrostatic equilibrium, or have enough gravity to pull it into an ellipsoidal shape.
  2. It needs to orbit the Sun and not any other body.
  3. And it needs to clear its orbit of any planetesimals or planetary competitors.

The planetary scientist's proposed geophysical definition for a planet:

A planet is a sub-stellar mass body that has never undergone nuclear fusion and that has sufficient self-gravitation to assume a spheroidal shape adequately described by a triaxial ellipsoid regardless of its orbital parameter.

I think the geophysical definition is the way to go, but we should expand it to include any hydrostatic equilibrium, not just a triaxial ellipsoid, including contact binaries, ternaries, quaternaries and even torus shapes if such exist.

So, let’s take a look at that now.


Incidentally, I think we should include rugby ball shaped planets like Haumea, and as well as that, contact binary planets too.

Artist's impression of Haumea spinning. Its dimensions are around 1920 x 1540 x 990 km - see The size and shape of the oblong dwarf planet Haumea

It has this rugby ball shape, not because of tidal effects of its moons, but because it spins so fast. The mathematician Jacobi predicted back in 1834 that sufficiently rapidly spinning planets would e"break symmetry" and become triaxial spheriods instead of oblate spheriods.

As it spins faster, other shapes become the preferred such as a contact binary. I think it’s entirely possible that we find a “Roche-world” type planet in our solar system, after all there are many contact binary asteroids and comets, so why not a contact binary planet?

Actually, a recent discovery, New Horizon’s next target 2014 MU69 for its extended mission (after Pluto) could possibly be a contact binary too, or possibly two very large bodies orbiting each other.

New Horizons' Next Target Just Got a Lot More Interesting This is the result of occultation experiments when 2014 MU69 passed in front of a star as seen from Patagonia, Argentina on July 2017. It could alternatively be an elongated body with a large chunk taken out of it. It’s no more than 20 miles (30 kilometers) long, or, if a binary, each about 9-12 miles (15-20 kilometers).

That’s quite small but a lot larger than Rosetta’s target Comet 67P/Churyumov–Gerasimenko which had two lobes of size 2.6 × 2.3 × 1.8 km and 2.6 × 2.3 × 1.8 km - Wikipedia

This is the observation data, showing that a binary body could match the data, though an irregular elongated shape with a chunk cut out if it would fit it too.

See also New Horizons' 2019 Target: A Binary Body? - Sky&Telescope

Alternative shapes - if ellipsoid, it is very elongated.

We even have a geologically complex binary asteroid 250 km in diameter (that’s half the diameter of Vesta), and it has a 12 km moon :)

Artist’s impression of the contact binary asteroid Hektor, 370 × 195 × 195 km with its 12 km moon. It’s the largest Jupiter Trojan, so it orbits the sun synchronously with Jupiter and is in its leading L4 point so orbits the sun ahead of Jupiter a sixth of the way ahead in its orbit. Discovered in 1907.

Artist’s impression of a cutaway of Hektor and it’s moon. The study suggests that it is porous with a denser core of compacted ices.

For more about this intriguing asteroid, see Scientists reveal puzzling orbit of Trojan asteroid's moon -

In total there are 62 known or suspected contact binary minor planets and 4 contact binary comets as of 15 July 2017. Of those, six have largest dimension greater than 100 km. See Contact binary asteroids and comets

With such a large contact binary asteroid as Hektor, so close to us, sharing Jupiter’s orbit - why not a “contact binary planet”? Like two Ceres sized asteroids / planets touching each other?

Now, if you just were to gently move two copies of Ceres to touch each other, they would of course coalesce under gravity to a single sphere shape. The result has to be rapidly spinning to be in equilibrium as a contact binary planet.

If Haumea was unique then I think the chance of a contact binary planet would be low.

But Haumea is just one representative of a whole class of bodies, the Haumea family, probably formed from break up of a larger body, most of which have high spin rates. Many of them have double peaked light curves - which suggest a triaxial ellipsoid rather than albedo variations. See Rotational properties of the Haumea family members.

The family includes the interesting:

  • 2001 QG298, a small Plutino (122 km in diameter), though it may be a binary with a very small separation. If it is dense, probably part of the proto Haumea core, then it is a contact binary. If it is less dense, made primarily of ice, it’s a triaxial ellipsoid like Haumea.

The Haumea family is thought to have formed as a result of collisions.

Also the Pluto system shows signs of collisions with its moons apparently contact binaries too, and perhaps even the entire system was a result of collision of two large bodies with the debris forming its moons. Indeed its moons two show signs of being contact binaries. It’s a little easier for such small bodies to be contact binaries as they don’t have to reach gravitational equilibrium.

However the main thing is that the Pluto system, which we have seen close up, gives good evidence that collisions are very common, or at least, they were back in the early solar system - even out as far from the sun as Pluto.

The New Horizons data suggests two, and possibly all four of Pluto’s smaller moons are results of mergers of smaller moons. See also Moons of Pluto - Wikipedia

If collisions are common, then mini planets rotating at high spin rates are possible. If that’s possible, then perhaps the Haumea family may be a representative of a large population in the Oort cloud.

If Haumea spun just a bit faster, as the result of a new impact, say, then the two hemispheres could pull apart into a contact binary. This has to happen if it spins fast enough to get a specific angular momentum (or angular momentum per mass) H > 0.308. This is the largest angular momentum for which a fluid ellipsoid is stable, see page 11 of this paper.

Asteroids that spin faster than this, if they are “loose rubble piles” are likely to pull apart into a contact binary. Finally, fission into two separate bodies happens when they reach H > 0.48412.

In the case of asteroids, they can spin up due to the Yarkovsky–O'Keefe–Radzievskii–Paddack effect - photons from the sun are absorbed and re-emitted later on as thermal radiation - or they are diffusely reflected and this tiny effect increases the spin rate, very slowly, but over time it all adds up. 54509 YORP will double its spin rate in 600,000 years. Eventually they spin so fast that they start to fly apart.

Then eventually those thrown out stones will coalesce to form a partner making it a contact binary asteroid.


The situation with contact binary asteroids with their rubble heap composition is not quite the same as for a fluid body because there aren’t any good examples of Jacobi ellipsoids like Haumea amongst the asteroids but when they spin fast enough they behave rather similarly to a fluid. See this paper for techy details.

So, though the mechanism for spinning up a dwarf planet would have to be different, if they can be spun fast enough, contact binary Ceres sized Roche world planets seem rather likely. Larger versions of the 2001 QG298 as a result of larger collisions, or as the result of multiple collisions.

Exoplanets could even be contact binary Roche world type Earth sized planets:

(Image NASA) 'Double Earths' Could Be Fun Exoplanets To Hunt For -- If They Exist

This is explored fictionally in Robert Forward's "Rocheworld" which is an "overcontact binary"


So, to venture a fun speculation, if there are Mars or Earth sized planets out in the Oort clouds, perhaps we can find similar contact binary worlds there too. Of course they couldn’t have oceans of water as they would be frozen solid - but what about oceans of neon (say)? It’s one of the most abundant gases in the solar system, fifth after hydrogen, helium, oxygen and carbon. So more abundant even than nitrogen, but it is very rare on Earth. It’s likely to be common further out in the outer solar system.

Neon is fluid at temperatures just a little too low for Pluto, at one time there was a speculation that Pluto might have neon rivers on it. It’s got a narrow range of only 24.55 °K to 27.05 °K (−248.45 °C to −245.95 °C, or −415.21 °F to −410.71 °F) to be liquid. Pluto’s temperature ranges from 33 °K (-240 °C -400 °F) and 55 °K (-218 °C or -360 °F). So Pluto is just a bit too warm for liquid neon.

Further out even Helium would be liquid. Could we have an Earth sized planet with a helium ocean?

Anyway fun speculation aside, if we find such a body, then so long as it is in hydrostatic equilibrium under gravity, I think it counts as a planet myself - a contact binary planet. You’d use the word “planet” in the singular there as you do for a contact binary asteroid.

Other shapes are possible as the planet spins faster, though it needs to be spinning so fast it is almost breaking apart before you can get them:

A rapidly spinning planet first becomes a dual lobed "Roche world" then finally three, and then four lobes become favoured, and a sufficiently rapidly spinning planet, just on the verge of breaking apart, is most stable as a torus, with a hole in the middle, like a doughnut. If you could somehow spin a planet so fast it is just about to break up, you might be able to trigger it to reform a donut shape. So could that happen in nature? For more on this see my So You Thought You Knew What Spinning Planets Look Like? ... Surprising Shapes Of Rapidly Spinning Planets

As that diagram shows, it’s even theoretically possible to have a contact ternary or quaternary planet and a donut shaped planet is a remote possibility though it wouldn’t be stable for long as its spin rate slowed down so I don’t think that’s too likely. But a rapidly spinning contact binary does seem possible.

If there are huge populations of thousands of contact binary planets, maybe we find a few contact ternary and quaternary and even a donut shaped mini planet. I don’t think that’s at all impossible.

There wouldn’t be much to slow down its spin rate out there and if the collision that formed it was recent, then it could still be in those shapes.


I know that the IAU definition is only intended for our solar system, and not for exoplanets. However if we do find non orbit clearing exoplanets, then we’d be in an uncomfortable position if we have planets that we would call dwarfs if they were in Earth’s system and non dwarfs around other stars. And then, as we get better at detecting planets, eventually surely we will find the equivalents of Ceres and Pluto in other star systems and then it would be natural to ask if we should apply the IAU definition there or not. .

Let's use Margot's π for our calculations here. The various measures are all are in pretty good agreement about where to draw the line between planets and non planets - and that's the only one that's easy to calculate. Very easy. It also doesn’t require you to know whether the planet has actually cleared its orbit or not. It’s a measure of its potential for orbit clearing and can be worked out using reasonably easily observed parameters that we can discover, even at a distance of light years, for many exoplanets.

Here is the formula:

With the planet m measured in Earth masses, M is measured in solar masses, and a in au. And k for those units is 833. 

The planet masses are often given in Jupiter masses. Jupiter is 317.83 Earth masses, and 833*317.83 =264752. So then the formula is

264752*m / ((M^(5/2)*( D ^(9/8))

m= planet mass in Jupiters, M= star mass in sun masses, D= distance in AU.

To make this number low, we want M and a large, and m small. So we want a heavy star with a small planet far from it. So, I sorted the Wikipedia list of exoplanets with the most massive star first.

One of the best candidates is Fomalhaut b. This one has actually been imaged :). Or at least a cloud or dust around it has been.

Fomalhaut b is in a 2,000 year highly elliptical orbit, from these observations by Hubble.

It orbits at a distance of 115 au around a star that has mass at least 1.9 times the mass of the sun. It is probably a planet of mass at most twice that of Jupiter though probably much less. So putting this into the equation (calculation indented):

M =1.95

m = 2

a = 177 ± 68

264752*2 / ((1.9^(5/2)*(109^(9/8) )

gives a Margot π of 543. So if that’s right, it’s comfortably within the range of the IAU planets.

It is "shrouded in dust but very plausibly a planet identified from direct imaging"

Artist's impression of Fomalhaut b (courtesy NASA). If it is a gas giant as shown, it will have Margot's π > 1.

However, it could be a much smaller planet as they don't know its size. What could its mass be to give a Margot π less than 1? Let's use Earth masses:

833*m / ((M^(5/2)*( D ^(9/8))<1

so m < ((M^(5/2)*( D ^(9/8)) /833 = ((1.9^(5/2)*(109^(9/8)) / 833 = 1.17.

So if the mass of Fomalhaut b is one Earth mass, it could have Margot π = 833 / ((1.95/2*(109 9/8) = 0.854

According to one theory from 2012 for Fomalhaut b it may consist of two shepherd planets which help constrain that ring, much like the way the shepherd moons Cordelia and Ophelia constrain Uranus’ narrow ring.

If so they get the narrowest ring if one of the planets is about the size of Mars and the other is about three times the size of Earth. In that theory, then the Mars sized planet would have Margot's π less than 1. The planets don’t need to be in resonance.

The brightest narrow ring of Uranus has two shepherd moons Ophelia and Cordelia. If Fomalhaut b is one of a pair of shepherd moons, the other one could easily be about the size of Mars. It could be the inner shepherd for Fomalhaut’s ring.

So we do already have at least one exoplanet theory with a non orbit clearing Mars sized exoplanet considered.

There are many other ideas for Fomalhaut b including a Jupiter sized object and a smaller planet with a ring system.

So - this doesn’t at all prove that we have found an IAU style non planet exoplanet. But it shows that such an object is sufficiently possible to have been considered as an astronomical hypothesis.


Then, what about a gas giant?

According to conventional received wisdom, it should be impossible for a planet to form at a great distance from its host star. Stars yes, brown dwarfs even, yes, as dual condensing clouds, forming in a similar way to binary stars. But what about a planet? Well the discovery of HD 106906 b shows that it is possible. It is around eleven times the mass of Jupiter, directly imaged, at a distance of 650 au from a star of mass about 1.5 that of the Sun.

So its Margot’s π is 723 (you can click on the link to see the calculation in google calculator). So this one does pass the test as a fully fledged exoplanet if we were to apply the IAU definition to other solar systems.

To be a dwarf exoplanet at that distance it would have to have mass 0.015 times the mass of Jupiter or less. So, an object with four times the mass of Earth at that distance would count as a dwarf exoplanet. And where there is one planet, perhaps there are more smaller ones?

At any rate, though not a dwarf exoplanets, it opens up the possibility of gas giants in distant orbits around other stars (using the IAU definition of a dwarf planet).


Let’s try a Saturn sized gas giant (0.299 Jupiter masses), and let’s try a heavy star of mass 4.5 times that of the sun, same as the middle of the range mass of HD 13189 with mass between 2 and 7 times that of the sun.

That’s the heaviest known star with a planet - then if it had a Saturn sized planet at a distance of 800 au, it would be just on the border and marginally not a planet according to Margot’s π = 0.998856

A Neptune sized gas giant (0.054 Jupiter masses) around such a star could orbit much closer and not be a planet.

Make it a star of six solar masses, within the possible mass range of HD 13189 and certainly a possible mass for a star with planets - then a Neptune sized gas giant could orbit at a distance of 100 au and still comfortably be a non planet with Margot’s π = 0.9117

Here is another one, 2MASS J2126-8140 which orbits its host star at over 4500 AU, its mass is around 13.3 Jupiter masses, and it is definitely a planet and not a brown dwarf. Its host star is TYC 9486-927-1 with a mass of 0.4 solar masses. So its Margot’s π = 2701.

Again it is an IAU style planet. But if it’s parent star had a mass of say 7 solar masses and if this planet was a little under half its size at 6 times the mass of Jupiter, it would be a dwarf exoplanet again, 0.95.

It was hard to recognize as an exoplanet rather than a rogue. It was found first as an apparently isolated rogue planet or brown dwarf, but then the astronomers spotted that it was moving in the same direction and at the same speed as a nearby star.

Lonely exoplanet orbits its star at greatest distance yet seen

If it was orbiting even further, over one light year away, like the Nemesis hypothesis for our sun, it would be far harder to link to its host star. At a distance of one light year, or 63,241.1 au like the Nemesis hypothesis for our own solar system, and if it orbited a heavier star, say 3 solar masses, its Margot’s π =0.897, just as is, with its mass of over 13 Jupiter masses. It would be a rather large Super-Jupiter gas giant (IAU) dwarf exoplanet.

I think it’s only a matter of time, myself, before we spot not just a Mars or Earth sized dwarf exoplanets - but Neptune, Saturn and even Jupiter sized and Super Jupiter gas giant dwarf exoplanets too.


The “Planet 9” hypothesis would just squeak in as an IAU planet.

But there are several other Planet X hypotheses that would not.

An Earth sized object could be a dwarf planet and not a planet if it was far away as 700 au, the predicted semi-major axis for "Planet Nine" if it exists. Now, it's true that they predict it to be Neptune sized, and at that mass, 17 times the mass of Earth, it would sneak in as a planet.

But one hypothesis is that "Planet 9" is not one planet but several planets in resonance, which according to some Spanish astronomers may explain the observations better.

Worlds Beyond Pluto --"There's Not Just Planet 9, But Several Planets Beyond"

In that case they could be of different sizes, and what's to stop some of them from being too small to be IAU planets, yet larger than Earth?

This would hardly be unprecedented. After all when Pluto was discovered, the astronomers were looking for a much larger planet, large enough to perturb the orbit of Neptune, but they just found the tiny Pluto.

Then there’s also the possibility of other planet X candidates smaller than Planet 9.


Here for instance is a paper from 2017, by Kathryn Volk and Renu Malhotra, building on earlier work by Lykawka&Mukai in 2008, suggesting the possibility of a Mars sized object at a distance of 65–80 au to explain the “Curiously warped mean plane of the Kuiper belt”.

That’s based on the observation that the Kuiper belt bodies are in a plane offset by about eight degrees from the “Invariable plane” of our planets. For non technical account of their research by the astrophysicist Brian Koberlein, see Goodbye Planet Nine, Hello Planet Ten

In the solar system, then you can forget the term for the solar mass as that’s 1, giving Margot’s π = 264752*0.00034 / (80^(9/8) ) = 0.65, so this candidate would be a Mars sized dwarf planet non planet if it exists and has an orbit at 80 au.

At the lower range of their suggested orbital semimajor axis, of 65 au, it would have Margot’s π = 0.821 so even there, it would be too small to count as a planet according to the IAU definition at least using Margot’s calculation. The other methods are similar but with something as close to 1 as this, it’s possible that you get disagreements about whether it is an IAU planet or not depending on which metric you favour.


For an Earth sized planet to fail to be orbit clearing, we need Margot’s π = 264752*0.00315 / (a^(9/8) ) <1

so a > (264752*0.00315) ^(8/9) or about 395 au.


Pluto’s mass: 1.31 x 10^22 kilograms

Jupiter’s mass: 1.898 × 10^27 kilograms

So Pluto is 0.000006902 Jupiter masses.

So, we need Margot’s π = 264752*0.00315 / (a^(9/8) ) <1

so a < (264752*0.000006902) ^(8/9) or about 1.7 au.

So, if we replaced Earth by an object the size of Pluto, it would count as a planet here according to the IAU definition.

However, it would be too small to be a planet at the distance of Mars, or in the asteroid belt. If we replaced Mars or Ceres by Pluto, it would not be a planet according to the IAU definition.


At a distance of 0.39 au, then we need 264752*m / (0.39^(9/8) ) > 1

So we need m > (0.39^(9/8) ) / 264752 = 0.0000013095 Jupiter masses. So m > 2.485431× 10^21 kilograms. Somewhere between the mass of Ceres (9.39×10^20 kg) and the mass of Pluto.

If it had the same density as Mercury, which has a diameter of 4,879 km, and a mass of 0.00017 Jupiter masses, it would have a diameter of 4,879* (0.0000013095/0.00017)^(1/3) or around 964 km. Almost identical to the diameter of Ceres of 950 km and 41% of the diameter of Pluto.

So if Mercury had the same density as it does now but was just a bit smaller than Ceres, it would not be orbit clearing.

So, imagine a situation like ours, with astronomers in another solar system, identical to ours in all respects but instead of Mercury we have a slightly larger than Ceres sized planet there which is still orbit clearing (just). They would probably have discovered their Mercury in their equivalent of our nineteenth century.

Then with their discovery of Pluto, it would already be more than double the size of the smallest planet. And their smallest planet would be orbit clearing like Mercury.

I can’t imagine an alternative IAU passing that resolution in this alternative solar system, which might actually exist as a real solar system out there somewhere - or one very like it.

So I think the IAU definition was accepted as a matter of a historical accident that Mercury happens to be larger than Pluto.


This is another issue with the IAU definition, but I think not a serious one. It’s based on a rather vague notion of “planetary clearing” - I mean vague in the sense that there’s a continuous metric rather than a discrete one - then you also have the potential problem of boundary planets that you don’t know how to categorize.

You do get that to some extent though with the geophysical definition too. Vesta is almost rounded and you could envision that we get close up observation of some asteroid or comet in between Vesta and Ceres in size and we don’t know whether to classify it as a planet or not according to the geophysical definition.

There are asteroids that are not quite in hydrostatic equilibrium and have many of the features of dwarf planets such as differentiated interior.

It’s the same with brown dwarfs there's a gray area between the smallest brown dwarf and the largest Jupiter sized object. Yet we continue to use the term. We are able to work with vague definitions like this.

So I don’t think this really counts against the IAU definition. But the boundary for the IAU definition is perhaps rather more arbitrary than the one for the geophysical definition.


Incidentally, I also think “Planet 9" is a bad name because – even if you go by the clearing orbits definition, it might not turn out to be a planet at all if it consists of several dwarf planet non planets. Also, if it is a planet, how do you know it is the ninth?

It’s so far away that there could be another IAU planet closer to the sun than it. For instance, Kathryn Volk and Renu Malhotra’s Mars sized mini planet could be closer.

We need a better name for it, though I don’t know how this can be accomplished with everyone used to calling it “Planet 9". I call it that too otherwise I’d not be understood.

"Planet X" is not exact enough as it would also apply to Nemesis, Tyche, Volk and Renu Malhotra’s planet, the idea of an extra planet to explain the Kuiper cliff, and indeed, even to Pluto itself, which was the original Planet X. When Lowell talked about Planet X, though he predicted a larger object, his speculations lead to Pluto being found. So I think you can arguably call Pluto “Planet X” too.

It would be great if it had a name already like Nemesis, Tyche etc, a planet doesn’t have to be proved to exist to be named, it seems. But sadly, not yet. So, there's no choice, we are stuck with this for now, unless someone comes up with another name that "sticks". But the name is a clumsy one that's also likely to be inaccurate - and since when did we refer to planets by numbers anyway? Earth is not "planet 3" and Mars is not "planet 4".

It will get even more confusing if we do find a closer Mars sized planet and give it some other name and then continue searching for “Planet 9”, at that point I think surely it would need a new name. But we are stuck with it for now.

See also my “Pluto – When Is A Dwarf Planet Not A Planet?”


Mike Brown himself predicts using a statistical argument from the orbit of Sedna, that we are likely to find planets beyond Neptune that are Mercury, Mars or even Earth sized. He then goes on to speculate about the future IAU meeting after astronomers find a Mars sized Dwarf Planet Non Planet according to the IAU definition.
The video is here:

Paraphrase: "Sedna has a twelve thousand year orbit around the sun. And we happened to find this one almost at the closest point it ever gets to the sun. Not by coincidence, because there is only about a 200 year period, where we could have seen it."

"So, 200 years out of 12,000 years means a 1 in 60 chance of finding it. So either we are very lucky, or since we only had a 1 in 60 chance of finding it, probably there are 60 of them and we just found the one that happened to be close. 

"Maybe it's not 60. Maybe it's 30 and we got a little bit lucky.  Maybe there are 90 and we got a little bit unlucky. But there are a lot of objects in this very distant region where we never knew of anything before. 

"Now the fun thing to think about is, if there are 60 of these, about three quarters the size of Pluto, like Sedna, there are probably 

  •  30 objects the size of Pluto. 
  • 10 objects that are twice the size of Pluto,
  • Two or three objects that are three or four, and maybe even five times the size of Pluto in this region here.

"Our big goal now, and one of my current graduate student's PhD thesis is to find these objects. If there are some big objects, two or three or four times the size of Pluto, these things are the size of Mercury, these things are the size of Mars, these things are the size of the Earth. 

"I am willing to go out on a limb and say that we will find something like that the size of Mars somewhere in this region of space. And scientifically this is going to be fantastic because we are going to be able to get to learn about an entirely new class of objects and try to understand how they got there. But just as much fun, of course, is that this will cause astronomers to go into a tizzy again.

"Because, if you find it, what do you call it? [goes on to explain about dwarf planets and "dwarf planets" not being a planet]

"I actually believe that that's the right classification. Because I still think that this population deserves to be put together and that the planets are actually a special case. But I don't think most people are going to buy that.

"I think if you find something the size of Mars, something the size of the Earth, I think most people are going to want to call it a planet.If we are really lucky we will find them in two or three years. They have to be a little bit close. Otherwise it is going to take another ten or fifteen and some bigger telescopes. But that's where we are headed and that's where this whole field is I think going next. 


I think it is only a matter of time before we find dwarf exoplanet non exoplanets according to this definition, around other stars, whether or not we find them in our own solar system. And eventually we may find gas giant exoplanets and even Super Jupiter dwarf exoplanets. They might even orbit a light year away or more if there are Nemesis style planets too small to count as a brown dwarf.

If so, depending on the size of the host star, those would probably not be planets according to the IAU definition.

If that happens, I can’t imagine this idea being accepted as a way of classifying exoplanets.

If so, then our solar system would be the only one where planets are classified as such depending on whether they clear their orbit, and that only for as long as we don’t find a Mercury sized or larger dwarf planet non planet here as well. I think it probably will become increasingly awkward to have a definition like this that is only used in our own solar system even if we don’t find gas giant dwarf planets or Mars sized or larger dwarf planets in our solar system according to the IAU definition.

In short I think the IAU definition is going to be superseded at some point in the future. It’s got a “use before date”, the only thing is, we don’t know what that date is as it depends on future discoveries.

Why not overturn it already and just call Ceres, Pluto etc planets?

We can of course use orbit clearing as a way of classifying planets. Just not as part of the definition of a planet. I think that stretches our language too much.

Good definitions of our terms help us to think and reason clearly. It’s not for the sake of Pluto and Ceres. It’s for the sake of clarity of our own thinking, that I think we should call them planets.


As they put it in their poster

"As a geophysical definition, this does not fall under the domain of the IAU, and is an alternate and parallel definition that can be used by different scientists. It is "official" without IAU approval, partly via usage."

As examples of classroom usage in the poster, they give:

“In the 2020s, NASA will send a spacecraft to study the planet Europa, which orbits the planet Jupiter as one of its many moons.”

“Earth and Io are the only solar system planets where ongoing silicate volcanism has been directly imaged”.

“The aeolian planets of Venus, Earth, Mars and Titan have surfaces shaped by wind”.

I really like this approach. For too long the moons have been treated like the Cinderella’s of the solar system, not really proper planets. But if, for instance, Ganymede was swapped with Mercury, it would certainly count as a planet, being bigger than Mercury. I think it’s a rather artificial distinction to think that Mercury, as a planet, is more interesting than Ganymede because it happens to be able to clear its orbit. Both are interesting in different ways.

Pluto may be a planet again if these astronomers get their way — and Earth's moon, too


This is Emily Lakdawalla’s graphic of Every round object in the solar system under 10,000 kilometers in diameter, to scale - she is including rounded objects like Vesta as well as the ones that are pretty much totally in equilibrium. Many of them aren’t labeled of the dwarf trans Neptunian planets. There would be far too many to memorize them all. But who expects to memorize all the asteroids, comets, or indeed, nearby stars?

Who knows the names of more than a handful of asteroids?

Now I do think it is worth having distinguishing names to single out the larger terrestrial planets Mercury, Venus, Earth and Mars as one category and smaller planets like Pluto, Ceres etc as another. There the word dwarf is useful, as an adjective indicating the size of a planet.


So, I'd use the word "dwarf" to refer to the size of the planet, not to whether or not it clears its neighbourhood.

As it happens none of the dwarf planets in our solar system clear their neighbourhood, if you use the word to refer to e.g. planets of around the size of the Earth's Moon or smaller. I think that’s the only reason the IAU definition has had the amount of support it has had so far.

But if you think of “dwarf” as referring to size rather than this orbit clearing property, so long as you make it clear what you mean, you could call the Earth a dwarf planet. It’s a more natural distinction. Gas giants, and dwarf planets. You could call planets like ours “medium sized planets” but that’s a bit clumsy and compared to the gas giants they are tiny.

If you want a more precise term, Alan Stern suggested calling Earth, Mercury, Venus and Mars dwarf planets, and Pluto, Ceres, Eris etc "sub dwarf" planets. See Page on

Alan Stern's suggestion is to call Earth and Venus, Mars, Mercury - the planets to the left of the picture, dwarfs.

Then he would call the really tiny Pluto, Charon, KBO objects etc, to the right of the picture, sub dwarfs. In more detail, the planets are sub dwarfs up to 0.03 Earth masses, dwarfs up to ten Earth masses, sub giants up to a hundred, giants up to a thousand Earth masses, and above that they are supergiants. See his Table 2 here. That would make Jupiter at 317.8 Earth masses our only giant, Saturn, Neptune and Uranus sub giants and supergiants would all be more than three Jupiter masses in size.

But they are all planets. Image from: Illustrations - Roberto Ziche


If we go with this definition we end up with moons that are larger than planets. Indeed we get many “planet moons” all the way down to tiny Nereid.

But that's true already with the IAU definition, as Ganymede and Titan are larger than Mercury.

Pluto is about the same size as Neptune’s moon Triton.

Triton Vs. Pluto

The suggestion to call them “planet moons” fits that nicely. It makes sense from a geophysical point of view because they are objects large and diverse enough to count as planets in their own right, they just happen to be orbiting a larger planet such as Jupiter.

The largest moons of our solar system compared with the main planets, Pluto and Charon, which would both count as planets. Beyond that it begins to get a bit vague, Dione and Tethys are both pretty close to hydrostatic equilibrium, and as you go further to the right, Enceladus is pretty close too.

Even Nereid may be largely in equilibrium, a little smaller than Vesta.

Approaching Enceladus - though Enceladus is about the same diameter as Vesta, it’s much more rounded and I think though not totally in equilibrium, it’s pretty close and would surely count as a planet according to the geophysical definition.

So, a moon if it is in hydrostatic equilibrium, and orbits a planet and the barycenter is inside the planet it orbits (as it is with the Moon and the Earth), then it is a “planet moon”

In the case of Pluto and Charon, the barycenter - the point they both rotate around - is outside of Pluto, so although Charon is lighter than Pluto, it's not so clear whether Charon is orbiting Pluto or they both orbit the barycenter. So this is a double planet.

Like this: animation created before the New Horizons flyby of course:

See Can we call Pluto and Charon a "binary planet" yet?


Some of you may be one of the original 420 astronomers who voted for the IAU proposal, or may come to it from the point of view of someone who is mainly interested in the dynamical study of the evolution of the solar system and solar systems generally.

From that viewpoint it makes sense to have a name for objects that clear their orbit and ones that don't.

I think it's just a different way of slicing things up, and depending on your specialty and interests in astronomy it may be the more natural seeming division. Fair enough. I think the main thing that's confusing there is to use the word "planet" and especially the word "dwarf" to make that distinction. Because there is no historical or natural association of the way we use the words planet or dwarf in everyday use with the concept of orbit clearing.

You could say "I'll use planet like this" so forcing planetary geologists to come up with a new term to refer to the objects that they are interested in. But if that was the intent, it hasn’t worked. In practice what has happened is that there is one group of astronomers that follow the IAU definition and another that follow the geophysical definition.

If we find a planet that's non orbit clearing and as large as Mars or larger, I think the number of astronomers that follow the IAU definition is sure to dwindle. Not because there is anything wrong with singling out planet clearing as an interesting and useful concept, but because the language they use to describe it seems unnatural to many.

The way forward I think is to come up with a new word to describe “an object that clears its orbit” instead of trying to get the word “planet” to do double duty as a geophysical definition and a dynamical one.


We also already have at least one moon sized objects that is orbit clearing. Kepler 37b. Its parent star is 0.8 solar masses. Its radius is 0.354 times the radius of Earth. Mass around 0.01 that of Earth but there is a lot of uncertainty about its mass, especially it could be much heavier depending on what it is made up of. Using Margot's formula for Earth masses, 833*m / ((M^(5/2)*( D ^(9/8)) then it's 833*0.01 / ((0.8^(5/2)*( 0.1 ^(9/8)) = 194. 

So, it is well into the realm of a planet according to the IAU. And in that orbit, 0.1 au, it could still be orbit clearing with a mass of m =  ((0.8^(5/2)*( 0.1 ^(9/8))/833 or 0.000052 Earth masses. Around 3 * 10^20 kilograms. Less mass than Vesta.

So as our searches get more sensitive we will surely find "planets" that are smaller than the asteroid Vesta which is not quite completely rounded. How small could they get?


There could be some circumstances where you get a really tiny object that still is able to clear its orbit. Take for instance 3 Juno which you couldn’t possibly say is rounded and so can’t count as a dwarf planet under the IAU definition.

It has a mass of 2.67 ×10^19 kg

So to convert that to Jupiter masses, the mass of Jupiter is 1.898 × 10^27 kg

So, it’s 0.000000014 Jupiter masses

Could it be orbit clearing if placed in a small enough planetary system? Well if extend planetary systems to “sub brown dwarf”s then yes:

Cha 110913-773444 is either a “sub brown dwarf” or a rogue planet, either way, it seems to be forming a planetary system right now, so taking birth like a star, but so small it’s more like a rogue planet.

With a sub brown dwarf to orbit, like that, 8 times the mass of Jupiter, which in turn is 0.0009546 solar masses, then 61 Juno would be orbit clearing right out to 100 au! With Margot’s π = 4.09

What about a proper star? A red dwarf?

Put our Juno clone in orbit around the smallest red dwarf star EBLM J0555-57Ab with mass of 85.2 times that of Jupiter, then it’s still orbit clearing out to 1 au with Margot’s π = (264752*0.000000014) / (((85.2*0.0009546)^(5/2)*(1^(9/8) ) = 1.965

So we could easily detect non rounded planet clearing objects in exoplanet systems once we have methods sensitive enough to find them.


I think that also strongly suggests that it is best to just have orbit clearing as a separate concept from hydrostatic equilibrium and not to try to define a (non dwarf) planet as an orbit clearing subset of the objects in hydrostatic equilibrium.

I think it’s just a matter of time, though it may need James Webb and other future observatories, before we find:

  • Some planets as large as gas giants maybe even as large as Jupiter or larger that fail to be orbit clearing
  • Some objects too small to be in hydrostatic equilibrium that are orbit clearing (obviously will need very sensitive observations to detect these)

The only connection here seems to be a historical accident that in our solar system, objects smaller than Mercury all happen to be non orbit clearing - so far.


It doesn’t seem likely that it would. A petition signed by over 6,000 signatories (now closed) was delivered to the IAU after the Pluto flyby.

Petition graphic

That was an ideal time to change it, with Pluto in the news with its complex geology and close up images showing its vast planetary size. So I don’t think they will change short of something dramatic forcing a change.

But we can call Pluto a planet, and indeed our Moon too. The IAU don’t even have authority over peer reviewed papers with many scientists continuing to call Pluto a planet in their papers, never mind the general public. So if you want to call Pluto a planet - and Ceres too, just go ahead and do so! Indeed you can also call our Moon, or Europa or Enceladus, or whatever, a moon planet, or satellite planet.

So if you want to do this, just go ahead and do so! If anyone asks you why you call it a planet, you can say you are using the geophysical definition of a planet as recommended by many planetary scientists.

One of the great advantages of this is that we can also get to call our Moon a planet!


Our Moon would count as The Eighth Continent the second largest after Asia and five times the size of Australia. We now know of ice in permanently shadowed craters at its poles, just next to its peaks of almost eternal sunlight. The lunar ice has been stable not just for millions, but for billions of years. Meteorites there could come from early Mars, early Venus, and right back to the earliest stages of life on Earth itself, probably with uncontaminated pristine organics preserved for all that time, for us to analyse.

It's even still geologically active, with recently formed small cliffs, trenches (formed as recently as 50 million years ago) and enigmatic almost crater free patches of strangely patterned terrain like the Ina depression that formed on its surface, some as recently as fifty million years ago, There have been many surprises, as I explored in the Moon science surprises section of my Case for Moon First

Here are a couple of videos giving an overview of our neighbouring "Moon planet" according to the geophysical definition of a planet (paper here).

Maybe it looks a bit less interesting, because it is gray in colour instead of the reddish brown of Mars, and the black sky as seen by the astronauts? If you add in a blue sky with, suddenly the planet like status of the Moon becomes more obvious:

Exploring our Moon planet Original here Apollo 17 at Shorty Crater - blue sky from here

There I have done no editing at all, not even colour balancing, just replaced the black sky by a sky from Earth. For more examples see my What if the Moon had blue skies in my Why Humans on Mars Right Now are Bad for Science. Includes: Astronaut gardener on the Moon

Words can sometimes be very powerful in their effects. If we were used to calling our Moon a “Moon planet” I wonder if the Apollo missions would have been more likely to continue beyond Apollo 17? Is the reason that we focus so much on Mars as a destination for humans, instead of the Moon, just because we call Mars a planet?


This is partly new material and it also uses some material from my earlier articles:

See also the Quora answers to: If Pluto is not a planet, what is it?


The Society of Unapologetic Pluto Huggers on Facebook has over 1,200 members, do join if you are keen on Pluto:

“A group for Pluto huggers and those who think the current IAU definition of a planet is too restrictive and needs to be adjusted in a reasonable, scientific way that makes more sense, enabling dwarf planets to be considered as much a planet as Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, and nefarious Neptune! Just kidding, Neptune! For a gas giant, you rock, dude!”

See also:


To get alerts of these posts join the facebook group: Robert Walker - Science20 Blog Alerts

And for those who would like to support my blogging, I've set up a Patreon page


My aim here is to get readers thinking. What are your thoughts? Do say in the comments. Also if you spot anything to correct here, however minor, do be sure to say. Thanks!